The 5-year outcome study of recurrent depression by Kupfer et al1 confirms the conclusion of the 1990 published 3-year outcome study by Frank et al in which they participated2 that found highdose antidepressant maintenance therapy was effective in preventing recurrence. The other conclusion drawn in this 1990 article, that there was "a modest prophylac
*IPT indicates interpersonal therapy. tic effect for monthly interpersonal therapy" is questionable. In that study, they attributed the differences in outcome to two factors: (1) imipramine hydrochloride (mainly) and (2) monthly interpersonal therapy (IPT) (a "modest effect").
We believe that a third factor could explain some of the results. Apart from imipramine, there were two variables, not one, that distinguished the non-IPT groups from the IPT groups: (1) lack of "maintenance IPT" and (2) severance of the relationship with the interpersonal therapist. This is a situation of loss that could cause depression, not because